When Absa's personal banking unit began incorporating gambling spend into its lending risk assessments, its researchers described the pattern as a “huge predictor of delinquency.” That finding, reported in FocusGN and Moneyweb, was not a policy statement or a press release. It was an empirical observation from one of South Africa’s largest credit books. The bank found that consumers with identifiable gambling spend in their statements defaulted at meaningfully higher rates than otherwise similar borrowers.
For credit providers who obtain bank statements as part of their Reg 23A process, this creates both a risk management opportunity and a compliance obligation. The signal is there. The question is whether you are reading it, and whether your affordability record shows that you did.
The Absa finding: what the signal looks like in practice
The delinquency correlation is not simply the presence of any gambling transaction. The risk profile sharpens around frequency and aggregate spend relative to income. A consumer who placed a single R150 bet on a rugby match in three months presents a different picture from one who made 22 transactions totalling R4,800 on Hollywoodbets and Betway across the same period.
What makes gambling spend a useful credit risk signal (rather than a moral judgment) is what it reveals about the consumer’s actual discretionary income position. Gambling spend is real outflow. It reduces the cash available to service debt. And unlike a structured debit order, it does not appear as a declared obligation on any application form. It is visible only in the bank statement.
Common gambling merchant identifiers in South African statements include Hollywoodbets, Betway, Sportingbet, 10Bet, Supabets, and casino ATM withdrawals at specific venues. Online gambling platforms often appear with payment processor codes (PayFast, PayD) that require transaction description matching. The NCR has confirmed gambling spend as a documented red flag in its affordability review guidance, and the signal is consistently treated as a behavioural indicator rather than a categorical disqualifier.
The empirical pattern that emerges from statement analysis: high-frequency, high-aggregate gamblers have thinner effective discretionary income than their gross income would suggest, and they are more likely to be managing financial stress through gambling than engaging recreationally.
What Regulation 23A says about gambling spend
Regulation 23A requires credit providers to assess a consumer’s “necessary expenses” as part of the affordability calculation. The regulation does not enumerate gambling as a listed expense category, but the practical implication is direct: high-frequency or high-aggregate gambling spend reduces available discretionary income in exactly the way a recurring debit order does.
If a consumer earns R22,000 per month, has declared obligations of R9,500, and appears to have R12,500 in discretionary income, that calculation changes materially if R3,200 of that R12,500 is going to gambling merchants each month. The stated discretionary income is R12,500. The observable discretionary income (after accounting for gambling outflows visible in the statement) is closer to R9,300. For a short-term loan with a monthly instalment of R1,800, the affordability picture is substantially different depending on which figure you use.
The NCR’s affordability review guidance has explicitly confirmed gambling as a red flag that credit providers should assess. The August 2025 draft amendments to Regulation 23A tighten the minimum expense norms requirement, making it more difficult to rely on consumer-declared living expenses that do not reflect actual spending patterns. A credit provider who obtains a statement showing significant gambling spend but uses only declared expense figures in the affordability calculation is running a process that the tightened framework specifically targets.
The reckless lending exposure when you ignore what the statement shows
NCA Section 81(2) requires credit providers to consider “all information reasonably available” in conducting an affordability assessment. This is the critical provision. If the bank statement was obtained (as required by Reg 23A for income and expense verification), then the gambling transactions in it are information that was reasonably available. Ignoring them is not a neutral act.
For the reckless lending exposure this creates, the case trajectory in 2025–2026 is instructive. Courts and the NCT have moved from examining whether an assessment was conducted to examining whether it was substantive. The March 2026 judgment on Section 81 (which the court characterised as requiring genuine assessment rather than procedural box-ticking) signals that the defence “we conducted an assessment based on declared figures” is weakening. Where a credit provider held information (in the form of a statement showing financial stress signals) and did not account for it, the substantive assessment standard has not been met.
DCASA’s reckless credit process is the most direct route for consumers to challenge a credit agreement. In a reckless credit investigation, the debt counsellor reconstructs the consumer’s financial position at origination. If a bank statement from the origination period contains gambling spend that the credit provider should have factored into the affordability calculation, that statement is admissible. CDH’s November 2023 alert on credit provider affordability obligations described “disregarding information indicating the consumer was ignorant regarding risks” as a reckless lending ground. The same principle applies to disregarding information indicating the consumer was financially stressed.
A credit provider whose affordability assessment record does not address gambling spend visible in the consumer’s statement has a gap that a debt counsellor can use. The parallel to loan stacking is direct: both signals appear in the statement before they surface as bureau defaults, and both create exposure for credit providers who fail to account for them at origination.
The CDH November 2023 standard: what “available information” means
Cliffe Dekker Hofmeyr’s November 2023 alert on credit provider affordability obligations set out the clearest practitioner-facing analysis of the “all information reasonably available” standard. CDH noted that credit providers have a positive obligation not merely to collect documents, but to engage with the information those documents contain. Collecting a bank statement and filing it without reviewing it for financial stress indicators does not satisfy the Section 81(2) obligation.
The Nedbank 2025 case reinforces this through a different fact pattern. In that case, the consumer’s employment was on a 10-month contract, a precariousness visible in payslip data and potentially in income patterns on the statement. The court found that relying on gross income without examining employment stability was insufficient. The principle that runs through both cases: credit providers are expected to engage substantively with information that was available at origination, not merely collect it.
Applied to gambling spend, the implication is clear. If the bank statement shows significant gambling transactions, the assessment record must address them. Not in moral terms, but in affordability terms: what is the actual impact on discretionary income, and does the consumer still qualify after that adjustment?
How to systematically assess gambling in a bank statement
The assessment process needs to be consistent and documented. Inconsistency creates its own risk: if the credit provider flags gambling for some applicants and ignores it for others applying for identical products, the process is vulnerable to a challenge that the inconsistency was discriminatory or arbitrary.
A practical framework for gambling assessment in bank statement review:
Step 1: Identification. Classify all gambling-related transactions using merchant category codes and transaction description matching. This includes: named betting platforms (Hollywoodbets, Betway, Sportingbet, Supabets, 10Bet), casino ATM withdrawals at named venues, and payment processor transactions where the payee description includes platform names. Online poker and fantasy sports platforms should be included where identifiable.
Step 2: Quantification. Calculate the aggregate gambling spend per month and across the review period. Calculate gambling spend as a percentage of gross monthly income and as a percentage of discretionary income (after declared obligations). Document the transaction count alongside the rand amount; frequency is a separate risk signal from aggregate spend.
Step 3: Classification. Apply a consistent classification framework. Occasional recreational spend (for example, a single transaction below R300 in a three-month period) warrants documentation but may not materially affect the affordability decision. Regular high-frequency spend, or any pattern where gambling aggregate exceeds a defined share of discretionary income (the precise threshold is a policy decision for the credit provider), should be treated as a material factor in the discretionary income calculation. Single large events (a R8,000 casino visit in one month with no prior pattern) warrant documentation and lender judgment.
Step 4: Adjustment. Include gambling spend in the affordability calculation as a recurring discretionary outflow where the pattern is regular. If the consumer gambles R2,500 per month consistently across three months, treat that as effectively a recurring obligation for affordability purposes. The resulting discretionary income figure is the one against which the proposed new obligation should be assessed.
This process applies regardless of whether the assessment is conducted manually or via an automated system. Behavioural credit scoring systems that extract and classify transaction data can apply this framework at scale and consistently, which matters when a credit provider is processing hundreds of applications per day.
What the assessment record must contain
A defensible affordability record on an application where gambling spend was identified must document the finding with enough specificity to reconstruct the decision. “Consumer gambles” is not sufficient. The record should contain:
- The aggregate gambling spend in rand for the review period and per month
- The number of transactions
- The gambling spend as a percentage of gross income or discretionary income
- The adjustment made to the discretionary income calculation as a result
- The affordability decision after that adjustment
A record written in plain language might read: “Bank statement review identified gambling transactions totalling R3,200 across 14 transactions in the three-month review period (Hollywoodbets, Betway). This represents 8.4% of gross monthly income. After subtracting declared obligations of R9,800 and gambling spend of R3,200 from gross income of R38,000, discretionary income available for the proposed obligation is R890. The proposed monthly instalment of R1,400 is not supportable. Application declined on affordability grounds.”
That record defends the decision to the consumer, to the NCR in an audit, and to the NCT if a reckless credit challenge is raised. It demonstrates that the credit provider considered the information that was available, applied it to the affordability calculation, and reached a reasoned outcome.
Where AffyScore is used in the assessment process, the decision pack automatically documents gambling spend as a named component of the behavioural score, including the aggregate spend figure, transaction count, share-of-wallet percentage, and its contribution to the overall risk signal. The plain-language narrative is generated alongside the score, so the defensible record described above is produced at origination without additional manual documentation effort.
